Autoimmune diseases are chronic, disabling conditions, and as a group they represent a leading cause of death among women younger than age 65 years and NIH estimates that 5% - 8% (14 - 22 million people) in the United States are affected by these diseases. Despite the burden of autoimmune diseases, there has been relatively little epidemiologic research into their etiology. Autoimmune diseases are more common in women than in men and in some of these diseases, such as systemic lupus erythematosus, more than 85% of patients are female. The female predominance of this disease is one of the reasons it has long been assumed that estrogen-exposures affect the risk of developing lupus. This project involves my work with lupus and other autoimmune diseases. The role of genetic susceptibility in lupus has been extensively examined, but the idea that lupus can be explained solely by genetics is not supported by recent studies. The concordance rate among monozygotic twins is currently thought to be 25% - 33%, substantially lower than the 75% figure derived from small studies subject to substantial selection bias published in the 1960s and 1970s. Except for one area on chromosome 1, there is little agreement between research groups involved in linkage studies concerning genomic regions that are likely to be involved in susceptibility to lupus. The relatively small degree of progress in genetic research, despite considerable funding and effort, points to a clear need to look beyond genetics in order to understand this and other autoimmune diseases. My work within the intramural program of NIEHS, and in collaboration with extramural researchers, represents a broader approach to studying lupus. It is an approach that is based on the often-cited idea that lupus involves a combination of, and interaction between, genetic and environmental factors. This work focuses on genetic factors and on measures of endogenous hormone exposure, exogenous sources of estrogen, and on environmental exposures that may affect the risk of lupus and of other autoimmune diseases. Importantly, I have used sound methodologic designs to develop the studies within this research program, including population-based approaches to identifying patients and relevant comparison groups. I have also developed innovative approaches to exposure assessment within the context of case-control studies. The specific studies that have been completed or are in currently underway within this project include: 1) The Carolina Lupus Study - the first large, population-based case-control study of genetic, hormonal, and environmental risk factors for lupus in the US 2) Dietary Risk Factors for lupus - a prospective analysis using data from the Nurses' Health Study (I and II) 3) Parity and Risk of Systemic Sclerosis - an analysis using a unique resource, the hospitalization and pregnancy registries covering the entire population of Sweden Results from the Carolina Lupus Study indicate that that hormonal factors (e.g., estrogen and prolactin exposure) that may be related to disease acceleration or progression in mice may not be related to the risk of developing lupus in humans, and variability in estrogen-related exposures do not affect the risk of developing lupus among women. These results raise questions about the applicability of these murine models for studying etiology in humans, and raise questions about the long-held assumption about the role of estrogen in this disease. The Carolina Lupus Study has also found relatively weak associations (odds ratios 1.5 to 2.0) with specific immune-related genes that had been hypothesized to affect risk of lupus (e.g., polymorphisms in IL-1, TNF-alpha and beta, and CTLA-4). Some associations were not seen across ethnic groups or in both sexes, raising additional questions about the generalizability of genetic data with respect to lupus. In contrast, specific occupational exposures (silica, pesticides, and mercury) were more strongly associated with lupus (OR > 3.0), and associations were seen in whites, blacks, men, and women. In addition, this is the first epidemiologic study to demonstrate the involvement of gene-environment interactions in human disease.